Project Summary Acquired drug resistance of cancer cells during cycles of chemotherapy and recovery is a major barrier against effective treatment of cancer patients. Although the time interval between chemotherapy doses is necessary for recovery of normal tissues such as bone marrow, it allows surviving cancer cells to develop adaptive responses such as activation of pro-survival pathways to escape drug-induced toxicity. Despite availability of gene mutations and deregulated signaling pathways as therapeutic targets, pre- clinical and clinical studies show that cancer cells activate compensatory signaling pathways, rendering standard and molecularly targeted therapies largely ineffective in treating cancer patients. According to genome scale analyses of human colon tumors by The Cancer Genome Atlas Network in 2012 and an international consortium in 2015, more than 45% of colon cancers contain frequent mutations in the RAS/RAF/MEK/ERK signaling pathway, making it a therapeutic target for specific molecular inhibitors of this pathway. However, the use of inhibitors of this pathway triggers feedback signaling through activation of other kinase pathways to overcome the therapy toxicity. Technological limitations to reproduce cycles of chemotherapy and recovery using physiologic tumor models hampers a mechanistic understanding of the interplay between different pro-survival pathways and development of treatment strategies to maintain drug sensitivity of colon cancer cells. We address this unmet need by micro-engineering 3D tumor models that reproduce key biologic properties of solid tumors and enable molecular analysis of drug resistance. This technology is based on an aqueous two-phase system. A drop of the denser aqueous phase containing cancer cells is robotically dispensed into each well of a microwell plate containing the second, immersion phase. The drop confines cells to self-assemble into a compact tumor spheroid. The ease of robotic generation, long-term maintenance, and in situ cyclic drug treatment and recovery of tumor spheroids in a high throughput setting provides an unprecedented opportunity to unravel molecular events underlying acquired drug resistance of colon cancer cells. We will harness this capability along with molecular techniques to quantitatively evaluate the effectiveness of various design-driven treatments against colon tumor spheroids. We will evaluate different treatment regimens such as combination treatments and sequential treatments with pairs of molecular inhibitors or molecular inhibitors and standard chemotherapeutics to identify strategies to effectively block tumor growth and avoid excessive toxicity that may be caused to patients by dual drug treatment. We will accomplish our goals through three specific aims: (i) Drug exposure-mediated activation of feedback signaling and resistance, (ii) Association of drug resistance with cancer stem cell phenotypes, and (iii) Treatment strategies to overcome drug resistance and stemness of colon cancer cells.